As technology progresses, geometries of integrated circuits become smaller and smaller. Increasing varieties and quantities of circuit elements are integrated into semiconductor dies to reduce sizes and costs of electronics equipment. Tuning and adaptive circuits are two types of circuits commonly used in electronics circuits. An example of a tuning circuit is a digitally-controlled oscillator (DCO) used in a frequency synthesizer. The DCO may use a resonant circuit having an inductance and a capacitance to establish an oscillator frequency for the DCO. The capacitance may be provided by a selectable capacitor bank 10 as illustrated in FIG. 1, according to the prior art.
The selectable capacitor bank 10 includes a first capacitive element C1, a second capacitive element C2, up to and including an Nth capacitive element CN, a first switch 12, a second switch 14, up to and including an Nth switch 16, and control circuitry 18. Each of the capacitive elements C1, C2, CN is coupled in series with each of the switches 12, 14, 16, respectively, and each capacitive element and switch series coupling is coupled between a first terminal FT and a second terminal ST. A control terminal of each of the switches 12, 14, 16 is coupled to and controlled by the control circuitry 18, which receives control information via a control interface CONT. The frequency of the DCO is tuned by controlling the capacitance of the selectable capacitor bank 10, based on the control information, by selecting the appropriate combination of capacitive elements C1, C2, CN to provide the desired capacitance. Thus, there is a need to integrate capacitive elements C1, C2, CN and switches 12, 14, 16 into a common semiconductor die.
Micro-electromechanical systems (MEMS) devices, such as MEMS switches, are often integrated into semiconductor dies. However, as geometries of integrated circuits become smaller and smaller, fabrication of MEMS devices may become increasingly problematic. FIG. 2 shows a MEMS switch 20, according to the prior art. The MEMS switch 20 includes a cantilever 22 having a movable contact 24, a fixed contact 26 electrically coupled to a first terminal 28, a second terminal 30 electrically coupled to the movable contact 24, and an actuator 32 electrically coupled to a control terminal 34. When the MEMS switch 20 is actuated by applying an actuation signal to the control terminal 34, the movable and fixed contacts 24, 26 come together, thereby closing the MEMS switch 20. When the MEMS switch 20 is closed, the touching movable and fixed contacts 24, 26 have some contact resistance; therefore, the MEMS switch 20 is called an ohmic MEMS switch. When the MEMS switch 20 is not actuated by removing the actuation signal from the control terminal 34, the movable and fixed contacts 24, 26 move apart, thereby opening the MEMS switch 20.
During actuation, if the geometries of the MEMS switch 20 are sufficiently small, there may be a risk of the cantilever 22 shorting to the actuator 32. Additionally, the actuator 32 may be formed using a first metal layer (not shown) and the cantilever 22 may be formed using a second metal layer (not shown). Therefore, when the second metal layer is etched to form the cantilever 22, the metal actuator 32 may be undercut. Thus, there is a need to prevent cantilever 22 to actuator 32 shorting and to prevent metal undercutting of the actuator 32 in a MEMS switch 20.